Study Guides (238,613)
Canada (115,253)
Criminology (336)
CRIM 220 (17)

CRIM220 Midterm to Final.docx

7 Pages
Unlock Document

Simon Fraser University
CRIM 220
William Glackman

CHAPTER8–UNOBSTRUCIVE&ARCHIVALMETHODS  Archival data = police/government statistical reports, secondary data archives  Contemporary products = newspapers, television, magazines, webpages  Historical materials = old newspaper reports, books, historical documents Unobtrusive Measures  Common problem in research: Reactivity  Unobtrusive Measures = help avoid reactivity, because the data is less influenced by the researcher: o 1) Data produced without thought that the ‗evidence‘ might someday be scrutinized by social scientists (eg dinosaurs) o 2) Data usually not collected until time has passed Physical Trace  1) Erosion = wearing away/removal of products because of our presence/activity o Rate of tile replacement in museum – index of exhibit popularity o Wear on library books are index of reading consumption o Shoe wear (eg see if kids play out more)  2) Accretion = addition/building up of materials because of our presence/activity o Analysis of contents of ‗art‘ – understand cultural attitudes and character o Nose prints (eg children nose on exhibit glass – see who is interested, what is interesting) o Garbage (eg classify people‘s garbage) o CSI/SOCO (eg how they solve crime and analyze data)  Advantage: o inconspicuous (not constructed – produced in a way not expected to be analyzed) o anonymous (no conflict with consent, cant identify particular footprints) o surreptitious (without alerting participants)  Disadvantage: o unclear representativeness, selective deposit, selective survival (eg criminal evidence – won‘t leave everything, and some disappear trace evidence) o privacy/confidentiality/lack of consent (most would not want someone to go through their garbage Archival Measures Records, databases, collections, speeches, photographs, newspapers, books, diaries, letters, internet  Advantages: o inconspicuous, surreptitious o anonymous and/or identified data o Allow longitudinal analysis = data already exist and have extensive time span cover, longitudinal analysis can be done now, no need to wait 40-100 years o Can go back and subject it to a greater/different scrutiny (eg unlike oral history if someone died, can‘t ask again) o Less Cost o Less influenced by reactivity  Disadvantages: o Common identifier problem – don‘t need real id – but link common items can trace u o Selective survival (what is available are not necessarily the most important/relevant data) o Selective deposit (*ucr crime rate example) Visual Data Archival visual data in magazines, television, films, documentaries, video sharing websites (YouTube), Tumblr, graffiti on subways  ICT development leading to new sources o Facebook, twitter o The ‗sign in‘ scam (want u create account, keep records, like you and what you search together, send you ads) o The ‗cloud‘ (keep info on server, not on specific computer)  Video o Recording real life events, detailed analysis, facial recognition (eg Vancouver riot) o No need to sign in (eg google- in idea that computer camera can recognize you when you use it) o Google glass o Documentary of historical changes (eg studying societal norms about female sexuality – study James Bond‘s girls)  Shortcomings o Authenticity/credibility – easily forged or manipulated (Photoshop) o Photo reflect the gender/class/race of photographer (eg Hope in the Shadow project- give disposable cameras to downtown east side and take photos from their point of view, not the privileged outsiders or mainstream media) o Performative element – people pose for camera, distorting reality, don‘t tell much about typical conditions Example: Constructing Crime Statistics  Observed score = True score + degree of error 1. A crime occurs  the true score  Does someone perceive as a crime  Does the person call the police ** unreported crime = Dark Figure of Crime  Does the police respond  Does the police write a report? 2. OR  Crime witness by police  Police perceive it was a crime  Decision to write report 3. Coding 4. Become crime statistic  the observed score  Uniform Crime Statistics – UCR 1962 o Standardize reporting –  Violent crime – one count  Property crime – one count  Multiple offences – only most serious offence o Problems:  Cant compare before 1962  Overstate the relative frequency of violent crimes relative to property crimes  Underestimate amount of crime  Operation Identification program o Reporting increase when program introduced, people more likely to call when they think they have a higher chance of recovering stolen good CHAPTER9-EXPERIMENT Requirements for Causation (John Stuart)  Temporal precedence: o cause must be before changes (eg light cant turn on before switch)  Association: o changes should be associated with subsequent result – should be RELIABLE - if don‘t switch, don‘t turn on; if switch, might not be every time but will be on ―more frequent than you would expect on the basis of chance alone‖  Elimination of rival plausible explanations: o must rule out the influence of all possible variables other than the experimental cause The Experiment  Experiment: Pretest  Treatment  Posttest o Eg Measure attitudes  expose to stimuli  measure attitudes again o This simple situations provides compliant with the for first two requirements for ―causation‖  Independent (treatment) variable = the variable we are trying to assess – exist independently of actual execution of study  Dependant (outcome) variable = the variable believed to be affected by the ―casual‖ variable – depends on other factors Internal Validity The extent to which differences BETWEEN GROUPS can be unambiguously attributed to the experimental treatment itself, rather than to other factors Threats to internal Validity  History: st nd o events that happened between 1 /2 measurement, that could affect results – or outside influence (see newspaper, learning ideas from class)  Maturation: o process within the people themselves that change over time (eg pill for children to walk – they learn to walk becuz they grow; give test for 8 hours, negative view because they are tired)  Testing: effects of ‗taking a test‘ on scores in second testing o Pretest sensitization: simply because you have taken a test, you may become more sensitive to the issues involved that you wouldn‘t have been otherwise – now you pay more attention to info in news o Practice effects: the more you take a test, the more you learn about the test/process, which affect results regardless of treatment  Regression (towards the mean): the tendency for extreme scores to move/regress closer to the mean on second testing o Extreme scores caused by random variation, which is less likely to be repeated o The more extreme the first score – the greater the propensity  Selection (ch.4): of participants  Instrumentation: (ch8) biased introduced by way you mechanically collect information Comparison/Control Group  Experiment Group: Pretest  Treatment  Posttest  Control Group: Pretest  No Treatment  Posttest (water surfing videos instead of immigration videos, the IV) o Doesn‘t have to be ‗do nothing‘, can be ‗best known treatment‘ or usual treatment Threats addressed:  History: both groups subjected to same historical factors  Maturation: control group will change at the same rate as experimental group (both experience same time delay)  Testing: same test to both groups  So: If Exp and Control group are equal to begin; both treated alike in every aspect except for the imposition of IV o  can say difference in two groups is cause by: the treatment Ensuring Initial Group Equivalence  Assumption of pretest equivalence = ―assume two groups are equal to being with‖  Not always possible: o If sample is an existing group – sample from‖ jail vs fines‖ - see which is more deterrent- but already differ on many attributes o Self-selection/volunteer selection – give driving course to those who wants—motivation! Perform better o Mandatory experimental group – give driving course to those who are sentenced by courts – already have bad skills! Poorer results  Complete equivalence is relatively impossible – but can be enhanced by: o Random assignment: CHANCE is sole determinant of which group each person will be in – so any differences in between is equally distributed across the groups, making both fairly equal  More successful with larger samples. (eg flip coin 10 times, could be all heads) small group size leads to uneven distribution of characteristics o Matching: good for smaller sample – instead of assume equivalence – we CREATE it  Require data ahead of time, understanding of ―relevant‖ variables.  Get matched pairs of individuals, randomly assign either to experimental and one to control  If equivalence is achieved, pretest is not needed to confirm it. External Validity Th
More Less

Related notes for CRIM 220

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.